Smart Fish Feeder Using Arduino Uno With Fuzzy Logic

Controller
st
1 Nisa Hanum Harani 3rd Anny Nurbasari
2ndAnggi Sholihatus Sadiah
Applied Bachelor Program of Applied Bachelor Program of Economics
Informatics Engineering Informatics Engineering Maranatha Khristian
Politeknik Pos Indonesia University
Politeknik Pos Indonesia
Bandung, Indonesia Bandung, Indonesia Bandung, Indonesia
[email protected] [email protected] [email protected]

Abstract—Smart Fish Feeder is an automatic fish feeder For the method of supporting the problem in this research,
that can be controlled using android smartphone. With this we purpose the optimization of fish feeding by using the
tool, fish owners will easily adjust the feeding schedule fuzzy logic controller. A way to control the provision of fish
according to the recommended feed dose, and provide feeding in this study is determined by the rules of fuzzy
aquarium cleaning scheduling. In designing a fish feeder, there logic controller that is using the parameters of temperature
are some criteria such as temperature and time interval of and turbidity of water which is an important aspect in fish
feeding. The data collection has been done by interviewing cultivation [7] [8].
ornamental fish traders. The data also has been obtained from
literature studies that support problem solving theory. The With these problems, there must be a solution to solve
Laravel framework is used to interpret the system while existing therefore, in this research entitled ”Smart Fish
Firebase as the Database Management System. The Android is Feeder Using Arduino Uno With Fuzzy Logic C0ntroller”,
as a front end that interacts directly with the user. Automatic this study aims to predict the depletion of fish feed with a
fish feeding systems are implemented using arduino micro- technological tool that is needed to grow at this time, so it
controller and prototype feeding devices. This research uses can be applied in everyday life . With the smart fish feeder,
Fuzzy Logic Controller method. With the creation of the the fish owner can easily set the feeding schedule according
prototype smart fish feeder the device functions well in terms to the recommended dose of feed. Simple but very effective,
of both controller and push data. The results of the calculation
with this tool is expected to be an alternative solution for
of the duration of fish feed using the Fuzzy Sugeno Algorithm
have been successfully applied to the smart fish feeder
people who like to keep goldfish without any feeling of
worry when leaving home in a long time. Framework
Keywords— Smart Fish Feeder, Arduino, Android, Firebase Laravel used to intercept systems with databases using
and Fuzzy Logic Firebase as Database Management System and Android as
front end in-teracting directly with user. Automatic fish
feeding systems are implemented using a microcontroller to
I. INTRODUCTION produce a prototype smart fish feeder to allow fish owners
The advancement in the field of electronics is now to schedule fish feeding schedules.
dominat-ing in science and technology [1]. With the
advances in technology and the improvement of people’s II. RELATED WORKS
living standards, [2] ornamental fish become an integral part
of daily life. One of the hobbies that the community of seeks Often the fish die because most of the food intake [4][5].
to keep is the fish in the aquarium [3],like a goldfish chef. People who have a hobby of fishing fish are worried about
Goldfish is one of the most popular ornamental fish. This the feed that should be done every day. With the smart fish
fish has a diverse body shape and has a variety of colors, feeder tool then the owner of the fish can easily adjust the
namely red, yellow, green, black and silvery. This is due to feeding schedule according to the recommended feed
the ease of care, feeding and so forth. Fish feeding is the dosage [6]. So a regular diet can help predict the depletion
main ingredient [4]. Fish will eat all the food given. Often of fish feed every week [9]. Timely feeding and water
the pet fish die because most of the food is eaten [5]. Fish turbidity parameters are important aspects of fish farming
feeding is generally done manually by sprinkling fish feed [7]. Temperature parameters are also important to know
into the pond and this work is done regularly every day. how much temperature and oxygen content is present in the
People who have a hobby of fishing fish are worried about aquarium [8]. The ideal temperature for fish farming is a
the feed that should be done every day. With the smart fish stable temperature in the range of 23-30 ° C and there is no
feeder tool then the owner of the fish can easily adjust the significant water temperature difference between day and
feeding schedule according to t he recommended feed night (no more than 5 °C). Fish are classified according to
dosage [6]. However, for people who have a fairly busy cold water habitats (below 20 ° C), and warm water (above
level of work, certainly feel a little difficulty when going to 20 °C). The tolerance of warm water fish to turbidity is
leave home moreover in a long time. With the least intensity higher (25 NTU) than cold water habitat fish (10 NTU). In
of time at home they have. The usual alternative is to ask for this condition the fish will give a maximum response when
help to others like close relatives or to neighbors. But this feeding. Besides that the fish’s immune system also works
can lead to new problems, such as nothing can be asked to optimally in these conditions. The innovation engine is
help keep the fish in the aquarium. deliberately used for fish farming especially for fish farming
with the enclosure system[10]. Smart Fish Feeder using a
microcontroller like Arduino as Microcontroller serves as
the main of the system [11]. The microcontroller is even
more than a microprocessor because it consists of ROM

Authorized licensed use limited to: Mapua University. Downloaded on March 06,2023 at 08:24:28 UTC from IEEE Xplore. Restrictions apply.
(Read-Only Memory), RAM (Read-Write Memory), [23] The fuzzy logic controller process can be seen as
multiple ports (input) and output [3]. Popularity has been Figure 1
driven by the simplicity of the use of Arduino and a large
number of sensors and libraries available to improve the
basic capabilities of this controller [12]. Arduino is a
platform for physical computing that is open source [13].
The Arduino programming environment facilitates
developers to manage, organize, upload, and simulate
programs [14]. Arduino is not just a development tool but it
is a combination of advanced hardware, programming
languages and Integrated Develop-ment Environment (IDE)
[15]. This appliance can be used for remote monitoring over
the internet [16]. Arduino Uno is a microcontroller board Fig. 1 Fuzzy Logic Controller
based on ATmega328, This board has 14 digital In Takagi-Sugeno neuro-fuzzy inference is presented in
input/output pins (where 6 pins can be used as PWM the form of functional relationships rather than in the form
output), 6 analog inputs, 16 MHz crystal oscillators, USB of output membership variables for fuzzy sets. Sugeno type
connection, reset button. [17] Research predicted the inference system use the output membership function which
depletion of fish feed using fuzzy logic controller method. is linear or constant [24]
The fuzzy logic process is fuzzification (crisp blurred input
value), rule evaluation (fuzzy control), and defuzzification The input stages will map sensors, such as switches to
(fuzzy output for crisp value). The purpose of the fuzzy membership functions. The fuzification process will
control system is to replace a human expert by a fuzzy rule- transform the input sensor into a fuzzy set. Fuzzy sets will be
based system [18]. The FLC can convert the linguistic mapped by the inference process which will be the basis of
control based on expert knowledge into an automatic control the decision [25]
strategy [19]. The Fuzzy System Center is the knowledge Figure 1 shows the basic elements of the FLC, consists of
base, formed by IF THEN rules of obscurity [20]. Fuzzy four main components: fuzzification, liguistic rules, fuzzy
rules based on IF-THEN rules[11]. inference, defuzzification.
The knowledge base is comprised of two components [12] 1) Fuzzification : A step to converting the parameter value
namely called fuzzy sets (data base) and fuzzy control rule from a set of input or input that is firm to be a fuzzy input
base. The concepts associated with fuzzy sets are used to by the membership function into a linguistic variable.
characterize fuzzy control rules and fuzzy data manipulation
in an FLC. Previous research that discusses the smart fish 2) Rule-Base (Linguistic Rules): Contains IF-THEN rules
feeder is different from the current research. This research is that are determined through fuzzy logic.
different related to research method used, not controlled 3) Formation of Fuzzy Set (Inference) : Fuzzy rules are a
using a smartphone (android), the parameter used. While combination of several events consisting of 3 variables with
this study aims to predict the depletion of fish feed, 3 fuzzy sets of each variable that will produce fuzzy rules
therefore for the method of supporting problem-solving called inference.
using fuzzy logic controller method. By using the fuzzy
4) Process of determining crip output (Defuzzyfication) :
logic controller method it will get the prediction of the end
The determination of output in the form of a strict constant
of the fish feed. Parameters used are turbidity of water, using a weighted average formula.
temperature and time interval. Framework Laravel used to
intercept systems with databases using Firebase as Database Figure 2 shows the system block diagram, the input from the
Management System (DBMS) and Android as front end system is the value of turbidity sensor and temperature, but
interacting directly with user. Automatic fish feeding the system can also run when getting a trigger from the user.
systems are implemented using a microcontroller to produce Arduino performs data collection, the output from Arduino
a prototype smart fish feeder to allow fish owners to will signal the module esp8266. When the signal is received
schedule fish feeding schedules it will activate a servo that will drop fish feed in the
aquarium. The values of turbidity and temperature have
been set with a fuzzy formula to automatically create a
III. METHODS control for dropping fish feed.
The principle of smart fish feeders is that all tasks are Setting Output
Point
carried out together in a short time, the feeding time Arduino Uno
ATmega328
Esp8266 module
Fish feed container

calculated in milliseconds. Measurements are given by

sensors that installed in aquariums, using a real time firebase
database so that the monitoring or controlling process does
not increase on the server. The microcontroller used is Temperatured Sensor
DS18B20
Turbidity Sensor
Ultrasonic Sensor

Arduino Uno.
Fig. 2 Block System
The method implemented is FLC, this method is used to
develop a controller that will determine optimization of
feed-ing given the temperature and turbidity. Fuzzy logic Figure 3 shows the hardware block diagram. Arduino Uno
microcontroller is used as a data acquisition and data
controller is a changing linguistic control strategies derived
from expert thinking becomes an automatic control strategy processing unit. The turbidity sensor serves to measure tur-
bidity in the water. Temperature sensor serves to measure
the temperature in the aquarium. Servo is used to drop fish

Authorized licensed use limited to: Mapua University. Downloaded on March 06,2023 at 08:24:28 UTC from IEEE Xplore. Restrictions apply.
feed in the aquarium and the esp8266 module as a
connecting microcontroller with the internet. Feeder tanks
are places for fish feed in the form of pellets placed. B. Fuzzyfization
Ultrasonic sensor to measure fish feed residue. Is a curve that shows the value of data input points into
its membership degree which has intervals 1- 0. Based on
Arduino Uno Servo
Modul esp8266
ATmega328 SG90
Feeder
tank
the fuzzy set that is formed then the membership function
that will be used to represent data is a triangular
membership function, using equations. The following
membership function input variables

sample
Ultrasonic
Turbidity Sensor Temperatured The membership function of the temperature variable can be
Sensor HC- Sensor
aquarium SR04
DS18B20 seen in Figure 5 and the turbidity variable membership
function can be seen in Figure 6

Fig. 3 Hardware Block Diagram

Figure 4 shows the workflow system , the user will select

the control menu, the system enters the activity control.
Users can choose control, if manually on then the system
will give a choice of feed dropping interval. When changing
the control menu to off, the fuzzy function will be executed
and the interval for dropping fish feed will be calculated
based on defuzzification value. The prototype of smart fish
feeder shows in Figure 5
Fig. 5 Temperature Variable Function(X Value)
Start The water temperature in the tank not only affects food
intake, but also feed conversion rates, because they affect
the physiology of fish. It is reflected in the growth of fish
which has a direct effect on the quality, quantity, and time
Controling
Activity of harvest in the aquaculture system.

off
Set read
Manual sensors

on

Fish feed
Select Interval
dropped

Fig. 6 Turbidity Variable Function (Y Value)

End
Research [10] shows a decrease in the survival rate of fish
egg embryos occurs when an increase in fine particles by 5-
Fig. 4 Flowmap
10% at spawning.
IV. EXPERIMENT AND RESULT
In this section, we use the controlled of inputs value is 20 C. Fuzzy Set
°C for temperatur and 26 % for turbidity. Fuzzy rules are a combination of several events. As has
been explained that there are 3 variables with 3 fuzzy sets of
A. Fuzzy Sets Input variables
each variable which will produce 9 fuzzy rules called
Table. 1 Fuzzy Sets for Input Variables inference.
No. Variable Fuzzy Set Domain
1 Temperature Low 0 - 20
Normal 20 - 30
High 24 - 35
2 Turbidity Clear 0 - 25
Pretty Clear 25 - 50
Turbid 30 - 80

In Table. 1 we set the input value by the membership

function into a linguistic variable.

Authorized licensed use limited to: Mapua University. Downloaded on March 06,2023 at 08:24:28 UTC from IEEE Xplore. Restrictions apply.
 z = 10 – 0,25 (10 – 2) = 8

[R3] = IF LOW Temperature and Turbidit y

TURBID THEN Long feeding duration
predikat = low (20); turbid (26)
=0;0
=0
z = 10 – 0 (10 – 2) = 0

Fig. 7 Output Variable (Z Value) [R4] = IF NORMAL Temperature and Turbidity

The fuzzy rules that will be created are 9 rules with the note CLEAR THEN Often feeding duration
that each rule that is formed including all variables using predikat = normal (20); clear (26)
equation below is the 9th rule of the fuzzy set:
= 0 ; 0,8
[R1] = IF LOW Temperature and Turbidity CLEAR =0
THEN Medium feeding duration z = 10 – 0 (10 – 2) = 0
[R2] = IF LOW Temperature and Turbidity PRETTY
CLEAR THEN Long feeding duration [R5] = IF NORMAL Temperature and Turbidity
[R3] = IF LOW Temperature and Turbidity TURBID PRETTY CLEAR THEN Medium feeding
THEN Long feeding duration duration
[R4] = IF NORMAL Temperature and Turbidity CLEAR predikat = normal (20); Pretty Clear
THEN Often feeding duration (26)
[R5] = IF NORMAL Temperature and Turbidity
=0;0
PRETTY CLEAR THEN Medium feeding duration
=0
[R6] = IF NORMAL Temperature and Turbidity
TURBID THEN Long feeding duration z = 10 – 0 (10 – 2) = 0
[R7] = IF HIGH TEMPERATURE and Turbidity CLEAR [R6] = IF NORMAL Temperature and Turbidity
THEN Often feeding duration TURBID THEN Long feeding duration
[R8] = IF HIGH TEMPERATURE and Turbidity
predikat = normal (20); turbid (26)
PRETTY CLEARITY THEN Often feeding duration
=0;0
[R9] = IF HIGH TEMPERATURE and Turbidity
=0
TURBID THEN Medium feeding duration
z = 10 – 0 (10 – 2) = 0
[R7] = IF HIGH TEMPERATURE and Turbidity
the degree of membership value of each variable in each set
CLEAR THEN Often feeding duration
using the following equation
predikat = high (20); clear (26)
Input Temperature = 20 = 0 ; 0,8
µlow [20]=1; µnormal [20]=0; µhigh [20]=0 =0
(1) z = 10 – 0 (10 – 2) = 0
Input Turbidity= 26 [R8] = IF HIGH TEMPERATURE and Turbidity
µClear [26]= 0,8; µPretty Clear [26]=0,25; µturbid PRETTY CLEARITY THEN Often
[26]=0 (2) feeding duration
predikat = high (20); Pretty Clear(26)
after the value of the predicate is obtained then look for = 0 ; 0,25
the z value using the following equation: =0
z = 10 – 0 (10 – 2) = 0
[R1] = IF LOW Temperature and Turbidity
CLEAR THEN Medium feeding duration [R9] = IF HIGH TEMPERATURE and Turbidity
predikat = low (20); clear (26) TURBID THEN Medium feeding duration
= 1 ; 0,8 predikat = high (20); turbid (26)(115);
= 0,8 =0;0
=0
z = 10 – 0,8 (10 – 2) = 3,6 z = 10 – 0 (10 – 2) = 0
[R2] = IF LOW Temperature and Turbidity
PRETTY CLEAR THEN Long feeding duration D. Defuzzyfication
predikat = low (20); Pretty Clear(26) The values that have been obtained are used to determine
= 1 ; 0,25 the output, which is defuzzyfication. The crip output
obtained is the final score calculated. Because a non-zero
= 0,25

Authorized licensed use limited to: Mapua University. Downloaded on March 06,2023 at 08:24:28 UTC from IEEE Xplore. Restrictions apply.
predicate is only found in the rules: [R1] and [R2], then it is V. CONSLUSION
calculated only [R1] and [R2]. To determine defuzzification From the results of designing and testing this feed
are as follows: optimization tool, there are several things that can be
Z = ( predikat * Z )/( predikat ) concluded. There are limitation such as too long a delay in
Z = ((0,8x3,6 )+(0,25x8 ))/(0,8+0,25) the controller due to RAM limitations. Arduino UNO is not
= (2,88+2)/1,05 compatible with smart fish feeders because it is associated
= 3,25 with limited memory. An unstable internet connection affects
The defuzzification result of the duration of fish feeding was the response time of the device. The method used in this
3.25 research is the fuzzy logic controller method. The
application of the Fuzzy Logic Controller method on the
E. Result smart fish feeder prototype can serve to optimize fish feed
Then the results of the calculation of the duration of fish based on temperature and turbidity.
feeding using the Fuzzy Sugeno Algorithm from the
input above, the calculation results with a score of 3.25 VI. REFERENCE
with the output being the Medium Duration. [1] D. Wang, J. Zhao, L. Huang, and D. Xu, “Design of a smart
monitoring and control system for aquaponics based on open
wrt,” in 5th International Conference on Information
Engineering for Mechanics and Materials, 2015.
[2] D. C. Love, J. P. Fry, L. Genello, E. S. Hill, J. A. Frederick, X.
Li,
and K. Semmens, “An international survey of aquaponics
practitioners,” PloS one, vol. 9, no. 7, p. e102662, 2014.
[3] R. Suharmon and A. Bahriun, “Perancangan alat pemberi makan
ikan otomatis dan pemantau keadaan akuarium berbasis
mikrokontroler atmega8535,” Singuda ENSIKOM, vol. 7, no. 1,
pp. 49–54, 2014.
Fig. 8 Smart Fish Feeder [4] M. Sidik, T. Negara, and T. Zuraiyah, “Model otomatisasi alat
pemberian pakan pada ikan berbasis sms gateway.”
In this study the experiment was carried out by [5] M. N. Uddin, M. Rashid, M. Mostafa et al., “Development of
experimenting on / off status on the controlling menu on automatic fish feeder,” Global Journal of Research In
Android and then seeing how the response of the tool and Engineering, 2016.
push the data carried out by the microcontroller to the [6] H. Effendy, “Rancang bangun alat pemberi pakan ikan dan
server.
pendeteksi suhu air aquarium otomatis berbasis mikrokontroler
arduino uno,” MASITIKA, vol. 2, 2018.
Table. 2 Experiment Result
[7] R. Oktafiadi, “Sistem pemantau kekeruhan air dan pemberi
Fuzzy Response
makan
Temperature Turdibity Fuzzy otomatis pada ikan berbasis mikrokontroler,” Jurnal Ilmiah
No Weeks Teknik
(C) (NTU) Hour Second Elektro Komputer dan Informatika, vol. 2, no. 1, pp. 7–16,
2016.
1 First Week 20 20 4 14400
[8] M. Mukhsin, “Simulasi alat pemberi pakan dan pengendali
2 Second 20 25 4 14400 kincir air yang berdasarkan suhu dan kadar oksigen pada kolam
Week ikan gurami berbasis mcu at89c51,” Widya Teknika, vol. 18, no.
3 Third 20 35 5.3 19200 1, 2010.
Week [9] K. Premalatha, P. Maithili, and J. Nandhini, “Smart automatic
4 Fourth 20 40 8 28800
fish
Week
feeder,” 2017.
5 Fifth Week 20 45 16 57600
[10] H. C. Wei, S. Salleh, A. M. Ezree, I. Zaman, M. Hatta, B. M.
6 Sixth 21 20 4.5 16200 Zain,
Week S. Mahzan, M. Rahman, and W. Mahmud, “Improvement of
7 Seven 22 20 5 18000 automatic fish feeder machine design,” in Journal of Physics:
Week
Conference Series, vol. 914, no. 1. IOP Publishing, 2017, p.
8 Eighth 23 20 5.5 19800
012041.
Week
9 Nine Week 24 20 6 21600 [11] E. Haryanto, “Perancangan dan implementasi alat pemberi
makan ikan otomatis berbasis mikrokontroler at89s52,” 2014.
10 Tenth 25 20 7.6 27360 [12] G. Barbon, M. Margolis, F. Palumbo, F. Raimondi, and N.
Week Weldin, “Taking arduino to the internet of things: the asip
programming model,” Computer Communications, vol. 89, pp.
128–140, 2016.

Authorized licensed use limited to: Mapua University. Downloaded on March 06,2023 at 08:24:28 UTC from IEEE Xplore. Restrictions apply.
 [13] S. Arduino, “Arduino,” Arduino LLC, 2015.
[14] M. Soliman, T. Abiodun, T. Hamouda, J. Zhou, and C.-H. Lung,
“Smart home: Integrating internet of things with web services
and cloud computing,” in Cloud Computing Technology and
Science (CloudCom),
2013 IEEE 5th International Conference on, vol. 2. IEEE, 2013,
pp.317–320.
[15] A. S. Lehman and J. Sanjaya, “Automatic fish feeder using
microcontroller,” in Seminar Nasional Teknologi Informasi
Komunikasi dan Industri, 2017, pp. 345–351
[16] G. D. Puspa, “Rancang bangun kontrol pemberi makan ikan di
dalam akuarium melalui short message service (sms) berbasis
arduino uno,”Ph.D. dissertation, Politeknik Negeri Padang,
2016.
[17] O. S. Abdi et al., “Prototipe sistem sirkulasi air tambak dan
pakan ikan otomatis berbasis arduino,” Ph.D. dissertation,
Universitas Muhammadiyah Surakarta, 2018.
[18] D. Bawa and C. Patil, “Fuzzy control based solar tracker using
Arduino uno,” International Journal of Engineering and
Innovative Technology (IJEIT), vol. 2, no. 12, pp. 179–187,
2013.
[19] Z. Roumila, D. Rekioua, and T. Rekioua, “Energy
management based fuzzy logic controller of hybrid system
wind/photovoltaic/diesel with storage battery,” International
Journal of Hydrogen Energy, vol. 42, no. 30, pp. 19 525–19
535, 2017.
[20] I. Ekpoudom, I. Archibong, and U. Itaketo, “Speed control of
dc motor on load using fuzzy logic controller (a case study of
emergency lube oil pump motor of h25 hitachi turbine
generator),” Nigerian Journal of Technology, vol. 36, no. 3, pp.
867–875, 2017.
[21] H. M. Hasanien and M. Matar, “A fuzzy logic controller for
autonomous operation of a voltage source converter-based
distributed generationsystem,” IEEE Transactions on Smart
grid, vol. 6, no. 1, pp. 158–165, 2015.
[22] K. Bendaoud, S. Krit, M. Kabrane, H. Ouadani, M. Elaskri, K.
Karimi,H. Elbousty, and L. Elmaimouni, “Fuzzy logic
controller (flc): Applicationto control dc-dc buck converter,” in
2017 International Conference on Engineering MIS (ICEMIS),
May 2017, pp. 1–5.
[23] “The car that adapts for you.”[Online].Available:
https://www.nxp.com/
[24] Y. Yudihartanti, “Analisis komparasi metode fuzzy mamdani
dan sugeno dalam penjadwalan mata kuliah,” PROGRESIF,
vol. 7, no. 2, 2015.
[25] C. C. Lee, “Fuzzy logic in control systems: fuzzy logic
controller. i,”IEEE Transactions on Systems, Man, and
Cybernetics, vol. 20, no. 2, pp. 404–418, March 199

Authorized licensed use limited to: Mapua University. Downloaded on March 06,2023 at 08:24:28 UTC from IEEE Xplore. Restrictions apply.